ArcFuels: Integrating Wildfire Models and Risk Analysis into Landscape Fuels Management

Joint Fire Science Program, May 12, 2009

Summary
That risk from wildfire continues to grow across the United States is not a new problem. Managing forest fuels in the real world—such as thinning and burning prescriptively—to reduce fuel loads have been used effectively to reduce the risk of severe wildfire. These actions have been helped by a variety of software tools that assist managers in planning and evaluating fuel treatments to ensure they are cost effective in terms of impeding the growth of future large, severe wildfires. While many landscape planning tools do a fine job within the scope of their capabilities, the process of fine tuning fuel management plans requires that users interact with large cumbersome databases and complex wildfire behavior models. The streamlined approach for modeling wildfire and planning fuel treatments on large landscapes developed in this study integrates fire behavior modeling and data processing tasks into a framework. This framework provides rapid assessment of wildfire risk and the potential effects of fuel management activities. The total picture of a particular scenario includes not only the predicted change in fire behavior, but also the change in likelihood of a fire, and resulting change in specific highly valued resources. Read further to learn about ArcFuels.

Key Findings
• ArcFuels, designed by the team, is a system that integrates a number of important fire behavior and vegetation models, geographic information systems, and desktop computer programs. It quickly and easily offers an approach for simulating, in real time, the effects of treatment plans.
• ArcFuels helps users enhance programs like FlamMap to calculate the potential effect of fuel treatments on burn probability and risk in terms of financial and ecologic value. This process offers a concrete measure of both wildfire benefits and damage that planners and landowners can use in fuels management plans.

Introduction
Where the sky is black and clean, and no urban lights pollute, you can see them. Bright points like stars—but moving—tracking steadily across the night. Wedded to us in orbit, they will emerge into day on the other side. They are satellites, looking at earth, acquiring data, sending us streams of information. Until relatively recently in the history of human instruments in space, many satellites were working efficiently, effectively, and independently. But understandings move forward, and thinkers and tinkerers developed a system to take advantage of the work the many satellites do, a system that connected the independent work being performed by diverse instruments and their individual systems. The global earth observation system of systems was born, an integration of efforts that can track weather events in real time, for example, and assist in predicting ramifications for people and lands in the path of a hurricane, or tornado, or storm.

The system of linked satellites that explores earth also helps scientists understand long-term weather trends, such as the melting ice sheets, rising sea levels, disappearing coastlines of our changing world. A user-friendly point of entry allows resource managers, decision makers, stakeholders to access the information that thousands of instruments produce in combined, understandable data sets. In less than fifty years, exploring earth moved from an era in which Shackleton’s three-masted, wooden-hulled barquentine sailed to Antartica, to man-made instruments (that record the break-up of polar ice sheets) roving through the scrim of the sky.

In no less a breath-taking leap, land management planning on federal lands in the United States moved from paper maps on drafting tables, and black and white aerial photos to the sophisticated geographic information systems and fire modeling tools that can simulate forest succession, fuel treatments and thousands of wildfires—in a matter of minutes. And all this with amazingly fine spatial detail. While many software programs used in landscape planning and fuels management work efficiently to solve small sequential steps in the planning process, someone forgot to link them together to help planners create the finished product—a fuels treatment plan that can withstand the test of National Environmental Policy Act (NEPA) and public scrutiny. A complicating factor was the difficulty of balancing the myriad goals that various land management agencies, local jurisdictions, and private and public owners hold. Looking at large diverse landscapes and evaluating the many alternatives for reducing wildfire risk to economic, ecologic and cultural values while garnering support from various stakeholders proved a problem. A system to streamline this process to integrate the many systems was needed. A solution to the process was proposed to the Joint Fire Science Program by Alan Ager, operations research analyst with the Western Wildlands Environmental Threat Center in Prineville, Oregon. With collaborators in research and on national forests, Ager and the team explored all the contents of a fire world.

Fire—In the tactile and worldly dimension
To begin effective planning, for any space, you need to know the boundaries of that space, whether it be the four walls of your home or the reaches of the entire planet. Like a watershed, a fireshed defines a certain space, a unit of land. The space embraced by a fireshed includes areas of land with similar fire regimes, fire history, wildfire risk, and potential for mitigation. Developed by the Forest Service’s Berni Bahro and his team of planners in the California region, the fireshed assessment process uses input from different stakeholders to simulate fuel treatments on the land, and to observe the resulting change in wildfire behavior in that space. Treated sites are placed to lessen the effect of wildfire on a fireshed, and are located strategically to block fire paths using ideas and software developed by Mark Finney at the Rocky Mountain Research Station. “The fireshed process was developed by the region 5 team as a mechanism to build consensus among landowners and concerned publics about wildfire issues and mitigation strategies,” Ager explains. This is especially important in the wildland-urban interface, where growing numbers of inhabitants, moving from urban centers, don’t always recognize that the greatest risks are from large fires that spread long distances to arrive at the boundary between wildlands and their real estate.

Burn models—Tinkering with the toolkit
To arrive at fireshed assessments, quick computing of multiple variables is critical. But problems occur when data sets can’t easily move among the fire behavior models, or vegetation and fuels programs, or geographic information systems, and even basic desktop office programs. Ager and his team created ArcFuels to eliminate the headaches of moving data from one process to the next. This system moves the data in the background, and helps planners and analysts organize the landscape and the planning process. The result is that users can leverage key fire models and visualization software to easily and handily design complex landscape treatment alternatives and test them in near real time.

In a collaborative setting, stakeholders and planners can quickly look at a range of thinning intensities in a specific overgrown ponderosa pine forest, or the effects of burning under different weather conditions, for example. Zooming out to the landscape, planners can test the net effect of a battery of stand treatments on the pace of a wildfire. Or the effect of omitting treatments on lands where the owner doesn’t want to allow these management tools. “ArcFuels,” Ager notes, “automates the process of scaling up individual stand prescriptions to a fireshed, and simulating the landscape package of treatments with wildfire simulation models.” ArcFuels organizes management prescriptions for stands in a project area within a geographic information system, he explains, and this simplifies the process of modeling all the complex concerns, constraints, conditions, management goals that multiple landowners in a fireshed may have.

At play in the field of chance
With the space defined as a fireshed, and software tools integrated to help with assessing fuel management actions, quantifying the change in risk from treatments remains. But how do you define risk, and how can it be calculated for highly random event like a wildfire? How do we place monetary value on things that hold intangible value? For example, while determining solid measures of harm to a human life may seem callous, it is the best we’ve arrived at when calculating insurance compensations and awarding personal injury damages. While acknowledging the intangible value, it is a means of applying a monetary value to calculate what could, or has been lost. Similarly, a forest, a landscape holds intangible as well as tangible values, and a loss or change in those values requires some way of calculating that change.

Building on previous papers by Mark Finney at the Missoula Fire lab, Ager devised a process to quickly calculate the expected net value change for many forest attributes like wildlife habitat, old growth, economic values and others. These calculations incorporate both likelihood, that is, the probability of fire at a specific intensity and location, and the net change in value as measured in financial or ecological terms. The expected net value change can include present and future, and positive and negative impacts from fire. Ager incorporated specific model linkages and code into the software to enable users to calculate the change in expected net value for fuel treatment scenarios. This achievement helped make it possible for fuel planners to use risk analysis in their fuels planning. In a central Oregon study of wildfire risk to northern spotted owl habitat, for example, the models were used to calculate the effects of fuel treatments on the probability of a fire with sufficient intensity to eliminate the key stand characteristics the endangered birds require.

By using ArcFuels to model different treatment options in real time, and see the possible outcomes of those treatments, managers and stakeholders can quickly find among many alternatives the best treatment placement and course of action. By using the risk framework, stakeholders can also quickly apprehend a hard sum of change in value associated with treatments. Like the satellites that can track that hurricane in real time today, and help in assessing the long term risks of more intense weather produced by a warming planet, the risk approach can show reluctant stakeholders the long-term hazards, with comparisons of suppressing wildfire or preventing wildfires through fuel treatment activities. With the virtual landscape shaped by using the tools, models and goals of wildfire risk, stakeholders can see a clear picture of what a “no action” decision really means in ten or in ten or twenty years. From sailing ships to space ships, from discrete specialties to integrated systems, our movements sometimes make giant leaps for mankind.

Management Implications
• ArcFuels is a new approach for melding the key technology ingredients for landscape fuels planning—geographic information systems, corporate databases, stand and landscape fire behavior models, and a streamlined process for developing and testing fuel treatment alternatives using risk-based measures. The system makes it possible for the first time to bring stakeholders and different land managers to the table to analyze fuel treatment scenarios in near real time, focusing the debate on the holistic and long term solution to the wildfire risk problem.

Further Information: Publications and Web Resources

ArcFuels website: http://www.fs.fed.us/wwetac/arcfuels

Ager, A.A. 2008. An ArcGIS interface to the Forest Vegetation Simulator for forest landscape modeling (in press) Proceedings of the third FVS conference proceedings, Fort Collins, CO. February 13–15, 2007.

Ager, A.A., M.A. Finney, B. Kerns, and H. Maffei. 2007. Modeling Wildfire Risk to Late Successional Forest Reserves in the Pacific Northwest, USA. Forest Ecology and Management 246:45-56.

Ager, A.A., M.A. Finney, and B. Bahro. 2006. Automating fireshed assessments and analyzing wildfire risk with ArcFuels. Forest Ecology and Management 234S:215.

Ager, A.A., M.A. Finney, and B. Bahro. 2006. Using ArcObjects for automating fireshed assessments and analyzing wildfire risk. Proceedings of the International ESRI Users Conference, San Diego, August 7–11, 2006. http://gis.esri.com/library/ userconf/proc06/papers/papers/pap_1547.pdf

Ager, A.A., M.A. Finney, and A. McMahan. 2006 A wildfire risk modeling system for evaluating landscape fuel treatment strategies. In: Andrews, P.L., Butler, B.W. (comp.), Fuels Management–How to Measure Success. USDA Forest Service, Rocky Mountain Research Station Proceedings RMRS-P-41. p 149-162.

Ager, A.A., A. McMahan, J. Barrett, and C. McHugh. 2006. A simulation study of forest restoration and fuels treatments on a wildland-urban interface. Landscape and Urban Planning 80:292-300.

Kerns, B. and A.A. Ager. 2007. Risk assessment for biodiversity in Pacific Northwest forests. Forest Ecology and Management 246:38-44.

Scientist Profile
Alan A. Ager is an Operations Research Analyst with the Western Wildlands Environmental Threat Center in Prineville Oregon. He received his Ph.D. at the University of Washington and has worked at the Forest Service for 20 years on a wide range of natural resource modeling problems.
Alan Ager can be reached at: Western Wildlands Environmental Threat Assessment Center, Pacific Northwest Research Station, 3160 NE 3rd Street, Prineville, OR 97754 Phone: 541-969-8683 Email: aager@fs.fed.us

Collaborators
Berni Bahro, Ken Wright, and Klaus Barber, USDA Forest Service Region 5
Mark Finney, Rob Seli, and Chuck McHugh, Missoula Fire Lab
Drew McMahan, Forest Health Enterprise Technology Team, Fort Collins, CO
Tom Burry and Bob Clements, Wallowa Whitman National Forest
Helen Maffei, Deschutes National Forest
Dave Owens, Ochoco National Forest
John Anderson, Balance Tech, Missoula, MT